THE EARTHQUAKE ADVANCE THEORY 171 change was as spasmodic as it was absolute, and we know of no better phrase with which to characterize it than to call it a glacier flood. Not only was the advance abrupt, but its termination was equally so, for it ceased with the season of 1906, and since then ablation has been at work healing the broken ice which has again relapsed into a stage of stagnation. The theory must, therefore, account for an abrupt and absolute cessation of the advance as well as for the spasmodic advance itself. Between 1899 and 1901, and also between 1906 and 1909, there have been slight fluctuations in the fronts of Turner and Hubbard Glaciers, which may be due to thrusts from tributaries; but there has been no notable advance of these great tidal, glaciers, although in 1909 Hubbard Glacier showed signs of the beginning of an advance which in 1909 we thought might prove to be of importance. Since, however, it was not followed in 1910 by a distinct forward movement we infer that it is not an advance comparable with those of the other advancing glaciers. Between 1906 and 1909 only one glacier, the Hidden, suffered advance, probably in the season of 1907. When last seen in early July, 1906, its terminus was in the same smooth, stagnant condition as in 1899 and 1905; but in early July, 1909, its front was about two miles farther down its valley. There had been accompanying spreading on the margins and a thickening of at least eleven hundred feet where the glacier terminus stood in 1905. With the advance came profound breaking of the surface, and, succeeding it, ablation in sufficient degree to render the surface passable. Thus the Hidden Glacier has repeated the phenomena of the other advancing glaciers, but the advance of its front has been greater than that of any other. Its change from stagnant to active condition and back again to stagnation was abrupt and all compassed in a short time, probably not over a year. In 1905 and 1906 Lucia Glacier showed no sign of past change, but in 1909 its moraine-covered lower portion, and also the valley glacier above, was greatly crevassed, there was a lateral spreading in progress, and a distributary on the western side had advanced notably, while the ice had surrounded and piled upon the northern end of the minatak near by. There was evidence that the advance and breaking had only recently begun and was still in progress. The surface was not so broken as the other glaciers, and the effect of the advance had not extended so far out in the stagnant piedmont bulb as in the case of the neighboring and contiguous Atrevida Glacier in 1906. Hubbard Glacier in 1909 showed an even earlier stage in advance than Lucia Glacier. Its eastern margin was newly broken, but what had occurred farther out in the glacier could not be told because the surface of this active glacier has been severely crevassed, except at the very margin, during the entire period of observation. However, the front of the glacier had moved forward somewhat since 1906 and there was evidence that the north arm had become more powerful. There was little additional advance in this arm between 1909 and 1910, though the west part of the glacier had continued the slower advance. In 1910 the Nunatak Glacier began to advance and up to June its tidal front had come out 700 to 1000 feet with some thickening, but, because confined in a narrow fiord, with practically no lateral spreading. This seems to have been the whole of the advance and is probably to be ascribed to a response to a thrust from a single small tributary.